Every idea about where the market gets props wrong, written as a fixed rule, recorded before the game and graded after. Wins and losses alike. Pick the ones that hold up.
Sport
Show
▼over the last▼
How to read this. The first test is the close: after a tracker flags a prop, did the price move its way before the game? Above 50% means the market tends to agree with it later, which settles much sooner than wins and losses do.
Trackers
Grading…
Build your own tracker
Write a rule in the Lab's formula scan, for example bsa >= 70 AND minutesTrend >= 10 or rippleDelta >= 3 AND hitSeason >= 60, check what it flags tonight, then press Publish. It joins this leaderboard as a community tracker, recorded and graded exactly like ours, and anyone can follow it. The rule is locked once published; unpublishing retires it, but its record stays up, so nobody can delete the losers and keep the winners.
Fields include bsa, hitL10, hitSeason, streak, l5, l10, season, line, price, oppRank, minutesTrend, gameEnv, rippleDelta, rippleGames, questionable, returning, clear, marketPct and more.
Every tracker records the side its rule picks, at that side's posted price (−250 to +250), the first time it flags a prop; later prices before the game are kept as the close. Players with too few games (8 in the NBA and NHL, 4 in college basketball, football and soccer) and yes/no markets are left out. Units are hypothetical, 1 unit a play. The no-skill baseline is what the same plays would return if the market's own odds were exactly right (the books' margin). Past results don't guarantee future ones. 21+.